Trugard and Webacy Develop AI Tool with 97% Success Rate Against Crypto Address Poisoning

Crypto cybersecurity firm Trugard and onchain trust protocol Webacy have developed an artificial intelligence-based system designed to detect and prevent crypto wallet address poisoning. This new tool is part of Webacy’s suite of crypto decisioning tools and utilizes a supervised machine learning model. The model is trained on live transaction data, combined with onchain analytics, feature engineering, and behavioral context.

The tool claims a success rate of 97% when tested against known attack cases. Address poisoning is a type of scam where attackers send small amounts of cryptocurrency from a wallet address that closely resembles the target’s real address. The goal is to trick the user into copying and reusing the attacker’s address in future transactions, resulting in lost funds. This technique exploits users' reliance on partial address matching or clipboard history when sending crypto.

According to a January 2025 study, over 270 million poisoning attempts occurred on BNB Chain and Ethereum between July 1, 2022, and June 30, 2024. Of those, 6,000 attempts were successful, leading to losses exceeding $83 million. This highlights the significant financial impact of address poisoning attacks in the crypto space.

Trugard’s chief technology officer, Jeremiah O’Connor, emphasized the team’s deep cybersecurity expertise from the Web2 world, which they have been applying to Web3 data since the early days of crypto. The team is leveraging its experience with algorithmic feature engineering from traditional systems to Web3. O’Connor noted that most existing Web3 attack detection systems rely on static rules or basic transaction filtering, which often fall behind evolving attacker tactics.

The newly developed system, however, leverages machine learning to create a dynamic system that learns and adapts to address poisoning attacks. O’Connor highlighted that the system’s emphasis on context and pattern recognition sets it apart. Webacy co-founder Maika Isogawa explained that AI can detect patterns often beyond the reach of human analysis, making it a powerful tool in combating address poisoning.

Trugard generated synthetic training data for the AI to simulate various attack patterns. The model was then trained through supervised learning, where it learns the relationship between inputs and outputs to predict the correct output for new, unseen inputs. This approach is commonly used in applications such as spam detection, image classification, and price prediction.

The model is continuously updated by training it on new data as new strategies emerge. O’Connor mentioned that a synthetic data generation layer has been built to continuously test the model against simulated poisoning scenarios. This has proven effective in helping the model generalize and stay robust over time, ensuring its efficacy in preventing address poisoning attacks.